Most patients with late-stage solid tumors are incurable with standard therapy. In addition, traditional treatment options often have serious side effects. Numerous attempts have been made to engage a patient's immune system for rejecting cancerous cells, an approach collectively referred to as cancer immunotherapy. However, several obstacles make it rather difficult to achieve clinical effectiveness. Although hundreds of so-called tumor antigens have been identified, these are often derived from self and thus can direct the cancer immunotherapy against healthy tissue, or are poorly immunogenic. Furthermore, cancer cells use multiple mechanisms to render themselves invisible or hostile to the initiation and propagation of an immune attack by cancer immunotherapies.
Recent developments using chimeric antigen receptor (CAR) modified autologous T-cell therapy, which relies on redirecting genetically engineered T-cells to a suitable cell-surface molecule on cancer cells, show promising results in harnessing the power of the immune system to treat B cell malignancies (see, e.g., Sadelain et al., Cancer Discovery 3:388-398 (2013)). The clinical results with CD-19-specific CAR T-cells (called CTL019) have shown complete remissions in patients suffering from chronic lymphocytic leukemia (CLL) as well as in childhood acute lymphoblastic leukemia (ALL) (see, e.g., Kalos et al., Sci Transl Med 3:95ra73 (2011), Porter et al., NEJM 365:725-733 (2011), Grupp et al., NEJM 368:1509-1518 (2013)). An alternative approach is the use of T-cell receptor (TCR) alpha and beta chains selected for a tumor-associated peptide antigen for genetically engineering autologous T-cells. These TCR chains will form complete TCR complexes and provide the T-cells with a TCR for a second defined specificity. Encouraging results were obtained with engineered autologous T-cells expressing NY-ESO-1-specific TCR alpha and beta chains in patients with synovial carcinoma.
Besides the ability of genetically modified T-cells expressing a CAR or a second TCR to recognize and destroy respective target cells in vitro/ex vivo, successful patient therapy with engineered T-cells requires the T-cells to be capable of strong activation, expansion, persistence over time, and, in case of relapsing disease, to enable a ‘memory’ response. High and manageable clinical efficacy of CAR T-cells is currently limited to BCMA- and CD-19-positive B cell malignancies and to NY-ESO-1-peptide expressing synovial sarcoma patients expressing HLA-A2. There is a clear need to improve genetically engineered T-cells to more broadly act against various human malignancies. Described herein are novel fusion proteins of TCR subunits, including CD3 epsilon, CD3 gamma and CD3 delta, and of TCR alpha and TCR beta chains with binding domains specific for cell surface antigens that have the potential to overcome limitations of existing approaches. Described herein are novel fusion proteins that more efficiently kill target cells than CARs, but release comparable or lower levels of pro-inflammatory cytokines. These fusion proteins and methods of their use represent an advantage for T-cell receptor (TCR) fusion proteins (TFPs) relative to CARs because elevated levels of these cytokines have been associated with dose-limiting toxicities for adoptive CAR-T therapies.